Intel 45nm New Stepping Q9550 C1

[AndreYang]

C1 looks good so far, but Q9550 price is really expensive.

I just test FSB 450*8.5 1.31V stable at 2XORTHOS Large. The FSB limit is about FSB 480 1M only.






QX9650 C1 will be coming soon.

[NeedMoMegaHurtZ]

Thx for the early results, Andre

So it looks like 500FSB might still be a dream even with the new stepping (at least with p35)

Do you happen to have any x48 mobo's in your arsenal? (Not that I'm expecting much more...)

[antari]

Hmm, non "(ES)" chip, these can't be far away! Nice result.

[okorn]

wow C1 seems to be pretty good

[jcool]

So C1 doesn't do wonders to the FSB wall
That's what really sucks with these smaller yorkies, they all hit the wall too soon, limited not by Mhz but by FSB. Still great for normal overclocks, but people looking for extreme results will have to stick with 65nm or pay top dollar for the QX.
That's why I decided to stick with my Q6600, I'll get that to approx. 4,2 stable under ss - can't do that with a Q9450

[CoolZone]

this is not even ES,I see it is retail or OEM?

[karl_eller]

Looking pretty good, but still not good enough. I was really hoping that they'd be able to hit about 500 FSB, otherwise I'll be looking at ditching my B3 for a G0 when the prices drop. If I get a decent batch I'll probably be able to get about 3.8-4.0Ghz on water.

Eller

[btdvox]

I have a feeling they compensated for something with C1 revisions for making them used with P35. Q9550 results were quite awesome when they first were leaked- Not saying yours are bad- But Q9550 were seen around 4.0.

They shouldve kept C0 and just make people upgrade there motherboard- If your going to buy a New proc thats this expensive same with the QX9650. I would hope you would have a new Motherboard with it.

[adamsleath]

still no volume of these little toys available for the pleb's.

[G80]

Will C1 be the first retail chips?. Because the first revision is nowhere to be seen.

[graysky]

I'm more interested in knowing if the thermal sensor fix has been implimented in these C1 stepping. Judging by the screenshot, the answer is yes.

@AndreYang - does coretemp go higher than 62 °C? Does the machine idle at reasonable temps as-per coretemp? Have you tried other temp monitoring software known to have problems with wolfdale/yorkfield cores like everest, hwmonitor, speedfan, etc.?

[Sav]

480 mhz fsb?

[eva2000]

oooh thanks for the heads up Andre - now to see if I can get my poor excuse of a QX9650 L739A644 retail oem (470FSB max quad 16m pi) exchanged for a C1 hehe

[MichelinGuy]

Intel said the first revision could be not stable, so that's why there's a C1 now. But I have to say that's a pretty nice OC for a quad at that voltage. Nice

[Giacomo]

The FSB limit is about FSB 480 1M only.

I'm waiting for my Q9450, and I realize that I won't see the nice 4.0 GHz number due to the impossibility to run 500MHz FSB 24:7 on a Quad.

About your 480 MHz limit... Do you mean that it's only good for 1M superpi, and no stable 24:7 system at all?

I'm going watercooling on MCH, MOSFET and CPU on my DFI X38-T2R, and I was hoping to hit at least a stable 470-480 MHz or so. What do you think about it?

Thank you!
Giacomo

[eva2000]

more likely 445-455FSB -/+ 10mhz 24/7

[pkrew]

A little disappointing, but it is only one cpu. Also only 1.31v was used for the tests and if only the 8.5 multi was tried we don't really know if the limitation was fsb or cpu clocks. Maybe increasing vcore and/or dropping the multi would have given more fsb. I just wish they'd release them for sale so we'd know for sure.

[Spyrus]

So it will do like Q6600 for 24/7 only with lower vcore and the benefits
are the extra cache and SSE 4.1.
Q9550@3.85 can be compared with a 4.0GHz Q6600.

[MarlboroMan]

seems like C-1 stepping likes FSB...
lets see 500 FSB

[jaredpace]

seems like C-1 stepping likes FSB...
lets see 500 FSB

AHAHAHA who is the guy on the right (eating vista) in this guys avatar???

[btdvox]

Intel said the first revision could be not stable, so that's why there's a C1 now. But I have to say that's a pretty nice OC for a quad at that voltage. Nice

Only because for some reasons there were users putting them on older P35 motherboards lol. Because of PSB noise from the cheaper 4 Layer PCB the 45nm C0 could not keep a stable fsb Voltage which would sometimes lead to instability.

Now why users would buy a 1000 dollar CPU and not have at least a X38 or a 780i - Beats me!

But for these new chips q9450, 9550 they're aimed at mainstream so not all users will upgrade there motherboard to compensate the new Chip.

But what sucks is there changing the QX9650 to a C1 also- i have a feeling this is going to be a bad idea for people wanting a QX9650 later on. I am sure that compensating for compatability with older motherboards will have an effect on how they OC.

QX9650's are a dream to OC I have mine running at 4.2 Stable with 1.38 Volts @ 66 C load under water

[MrRevhead]

Thanks for sharing Andre

Looks like the FSB limit on the Engineering Samples reflected those of your C1.

Pity, considering how these are scaling relative to VCore.

[Delirious.]

Does anyone have an idea when the QX9650 C1 will be available? I have searched and the best I could find was March 3. I have seen on another forum a C1 being tested but more than likely was released outside the US. I am not sure. Anyone know?

I am looking for the temperature readings to be fixed and better overclocks. I have one that I can take back as I am still in the window of returns. Anyone have a suggestion on what to do. Wait for the new ones or what?

[graysky]

C1 looks good so far, but Q9550 price is really expensive.

I just test FSB 450*8.5 1.31V stable at 2XORTHOS Large. The FSB limit is about FSB 480 1M only.

Is your Q9550 build still stable @ 8.5x450? I'm running my X3360 @ 8.5x400 stable, but haven't seriously tried higher. Can you share all your vcores so I can get an idea what range to set mine up to? Also, what are the hardware specifics for that system?

Thanks!

[lowtcell]

I can't get my Q9550 to boot past 435FSB. Not sure if the Asus P5E is limiting me or if the chip is just a dud.